1
00:00:00,000 --> 00:00:05,000
Thanks, much

2
00:00:05,000 --> 00:00:10,000
Hi everyone. My name is Chris Freelan and I'm a librarian at the Internet Archive.

3
00:00:10,000 --> 00:00:16,000
So here we are today at our final session of the library as laboratory series over the past 10 weeks 10 weeks.

4
00:00:16,000 --> 00:00:26,000
If you can believe it, we've brought together, some of the world's, leading scholars in the digital humanities and data intensive sciences to talk about their projects, and how they're using the Internet archive in the course, of

5
00:00:26,000 --> 00:00:34,000
their research, we started off our series by asking the bold question, What can you do with billions of archived web pages?

6
00:00:34,000 --> 00:00:43,000
And 2 weeks ago we ran it out. the formal part of our series by asking, What can you do with 60 million digitized pages of scientific literature in between?

7
00:00:43,000 --> 00:00:54,000
We've heard from bibliographers authors educators, archivists, and data scientists about how they're using material services and infrastructure from the Internet Archive.

8
00:00:54,000 --> 00:01:00,000
The recordings for all of those sessions are now available online. and I see the Duncan who's working behind the scenes today along with Caitlin.

9
00:01:00,000 --> 00:01:06,000
Hello, both of you, and thank you, has shared that out in the into the chat.

10
00:01:06,000 --> 00:01:10,000
But we're gonna do something a little different today when we started planning this series.

11
00:01:10,000 --> 00:01:21,000
We had inquiries from a number of scientists who wanted to do shorter presentations, you know, get something more like a a status update or an overview of the research, not a 45 min lecture, with

12
00:01:21,000 --> 00:01:34,000
a long Q. A. So we put out a call for lightning talks, And today we're gonna bring you a variety of short talks and videos from researchers working on topics as varied as understanding the effects of emotional

13
00:01:34,000 --> 00:01:37,000
queues by the chairs of the Us. Federal Reserve.

14
00:01:37,000 --> 00:01:44,000
On financial markets to a diy book scanning robot So here's the game plan for today.

15
00:01:44,000 --> 00:01:47,000
Live. transcripts are available, captions are available.

16
00:01:47,000 --> 00:01:51,000
Use the live transcripts feature of Zoom to turn those on.

17
00:01:51,000 --> 00:02:01,000
We also you can copy from within the within the chat by just mouseing over anything individual that you're interested like links that we're gonna share.

18
00:02:01,000 --> 00:02:09,000
And also we're gonna capture all of the links and the chat and make those available including the video, all the resources will be available.

19
00:02:09,000 --> 00:02:15,000
So if you're if you're trying to catch something if you want to remember something a copy a link that we've shared.

20
00:02:15,000 --> 00:02:28,000
Rest assured that will make it into the email that we're gonna send you tomorrow with with the video for today's session. and the other resources that we've shared as you see the chat is open please do

21
00:02:28,000 --> 00:02:33,000
be respectful and keep the comments on topic and use the Q. A.

22
00:02:33,000 --> 00:02:41,000
Feature to submit questions for our panelists and We'll have time for one, probably one question per speaker.

23
00:02:41,000 --> 00:02:53,000
So do please submit questions, for us to to gather through gather from, and a final thing I do want to mention about time. I anticipate that we're definitely gonna run over an hour today.

24
00:02:53,000 --> 00:03:06,000
Probably more towards 90 min. So for those of you who need to to to depart at the top of the hour. Rest assured we're gonna record all of this, and it'll be made available to everyone so for now I can see

25
00:03:06,000 --> 00:03:09,000
people are already using the chat. Please do say hello!

26
00:03:09,000 --> 00:03:14,000
Let us know who you are and where you're joining in from today.

27
00:03:14,000 --> 00:03:28,000
So I actually was just looking back over my my planning notes, and we started talking about what became this library as laboratory series last September, and, like many ideas at the Internet Archive, it burst forth with

28
00:03:28,000 --> 00:03:35,000
enthusiasm. and gusto in a meeting with Brewster Kale the founder and digital librarian of the Internet Archive.

29
00:03:35,000 --> 00:03:41,000
So i'd like to welcome brewster to the screen and Brewster, I'd like to have you share a bit of your thinking.

30
00:03:41,000 --> 00:03:48,000
Why was organizing this digital humanities expo a priority for you and for the Internet Archive?

31
00:03:48,000 --> 00:03:58,000
Oh, thank thank thank you. Chris. And the this is really fulfilling the dream of the Internet Archive.

32
00:03:58,000 --> 00:04:04,000
We We started by trying to archive the Internet and how do you go and do that, and then expand it to to archiving other things.

33
00:04:04,000 --> 00:04:08,000
But it wasn't just to make It so that you can go and find old web pages.

34
00:04:08,000 --> 00:04:19,000
It was to try to get a bigger view. Can we go and help make it so that lots of new and different things can happen without having to build your own collection.

35
00:04:19,000 --> 00:04:27,000
Yourself build a library that you can just go and have your new idea, and all of the materials needed for your research are on the shelves.

36
00:04:27,000 --> 00:04:34,000
But now just digital shelves. So that that was sort of the the impetus for the Internet archive in general.

37
00:04:34,000 --> 00:04:39,000
And so this idea of having people be able to use these collections at scale has been so important.

38
00:04:39,000 --> 00:04:44,000
But actually it's really pretty hard the collections are I think in pretty good shape.

39
00:04:44,000 --> 00:04:50,000
But they're just huge and hard to find and and figuring out how to go and and use them.

40
00:04:50,000 --> 00:04:56,000
But we have seen fabulous presenters, and including the ones that did today.

41
00:04:56,000 --> 00:05:05,000
But the urgency of Miss Information, and feeling that our information ecosystem is out of our control, I think we need.

42
00:05:05,000 --> 00:05:10,000
We really need people with a macroscope. This is Jesse Osbel's lion.

43
00:05:10,000 --> 00:05:18,000
He said. We got really far with a microscope towards under understanding, and humans and scientists and science got really further along.

44
00:05:18,000 --> 00:05:23,000
Now we need a macroscope How Can We kick a bigger picture of what's going on.

45
00:05:23,000 --> 00:05:29,000
Can we use web scale Tv scale, book scale types of information to try to help us understand?

46
00:05:29,000 --> 00:05:34,000
Our world and i'm jazz that many of the key players have come forward to speak about what they're doing.

47
00:05:34,000 --> 00:05:39,000
They're so inspiring and inspiring to me and then there's firing our staff.

48
00:05:39,000 --> 00:05:48,000
But I think others to go and say, yes, I can go and use these sorts of collections at scale. to go and do different things.

49
00:05:48,000 --> 00:05:53,000
So what's the value to the Internet archive of this just from a completely selfish point of view?

50
00:05:53,000 --> 00:05:56,000
It helps drive us forward to make more useful collections and tools.

51
00:05:56,000 --> 00:06:05,000
And so for that, we're gonna need your feedback let's do what the Internet archive can do to be a better library.

52
00:06:05,000 --> 00:06:11,000
What are the materials, the tools, the structures that you would like to see or platforms?

53
00:06:11,000 --> 00:06:20,000
Is this library as laboratory series useful to you we've seen a large number of people join in on these.

54
00:06:20,000 --> 00:06:27,000
So. but please feedback that that's absolutely essential and thank you, Chris, for making all of this.

55
00:06:27,000 --> 00:06:33,000
Come, come about with a tremendous group of of researchers that have come together.

56
00:06:33,000 --> 00:06:43,000
So looking forward to today. Thanks for that brewster always always good to to hear from you and hear what your what you're thinking about.

57
00:06:43,000 --> 00:06:49,000
So, as we jump in here today, I want to let all of you in the in the audience know what to expect.

58
00:06:49,000 --> 00:06:53,000
So we have 3 segments of talks and each segment will have 3 to 4 talks.

59
00:06:53,000 --> 00:07:03,000
If that's talks in videos so what we'll do is we'll move from talk to talk quickly, and we'll wrap up each segment with a quick round of questions and answers.

60
00:07:03,000 --> 00:07:06,000
So if you have questions again, drop them into the Q. A.

61
00:07:06,000 --> 00:07:11,000
So that we can we'll probably as I said have time for one question allowed.

62
00:07:11,000 --> 00:07:23,000
But our speakers are going to be hanging around, and so, if you drop off additional questions, they might be interested in engaging with you further and answering some of those questions from the from the Q.

63
00:07:23,000 --> 00:07:29,000
And A. So please as you have questions, please do ask them using the Q. A.

64
00:07:29,000 --> 00:07:42,000
Feature. i'll also mention that the duncan dropped a link into the into the chat with the agenda, so that you can follow along the the order that we're going in today so let's

65
00:07:42,000 --> 00:07:56,000
get started up first today. is Kate Milner from the University of Edinburgh Case gonna tell us about the forgotten histories of the mid-century coding boot camp over to Uk.

66
00:07:56,000 --> 00:08:07,000
Hi! everyone, Thanks so much for joining. Just gonna share my screen, and we will get this party started

67
00:08:07,000 --> 00:08:15,000
Okay, So, as Chris mentioned, i'm kate milner I am a Mari Curie postdoctoral fellow at the University of Edinburgh, and I'm.

68
00:08:15,000 --> 00:08:26,000
Just really grateful to Chris and the internet archive for having me to talk about some of the amazing artifacts that I've used in my historical research on electronic data processing schools which are what I consider to be

69
00:08:26,000 --> 00:08:32,000
the mid-century predecessor of the contemporary coding boot camp.

70
00:08:32,000 --> 00:08:46,000
So in the past decade new stories like these have become increasingly familiar. I'm sure most, if not all of you have seen or read an article that talks about learning to code. and what it can accomplish maybe you even wrote one

71
00:08:46,000 --> 00:08:49,000
of them. As a part of my PHD. thesis.

72
00:08:49,000 --> 00:08:55,000
I read over 200 articles over a 10 year period that talked about learning to code. In reading.

73
00:08:55,000 --> 00:09:08,000
All of these articles it became clear that coding has been positioned as a solution to a variety of interconnected social issues, including concerns about Ai gender and racial bias in the Tech sector economic

74
00:09:08,000 --> 00:09:17,000
inequality and skills gaps that have supposedly left millions of well-paid positions unfilled, due to a lack of appropriate training.

75
00:09:17,000 --> 00:09:24,000
In response to this discourse, an entire industry of coding boot camps has developed in the Us.

76
00:09:24,000 --> 00:09:32,000
And across the world. coding boot camps are short-term intensive courses that aim to make professional technologists out of technical novices.

77
00:09:32,000 --> 00:09:40,000
The topics of these pro programs can vary some are focused on data science, some on user experience, some on software engineering.

78
00:09:40,000 --> 00:09:45,000
There's even programs for digital marketing and product management the length of time can vary, too.

79
00:09:45,000 --> 00:09:59,000
Most are around 3 to 4 months, but some even last as long as 2 years. But independent of the topic and the timeline, the promises that these programs make are usually the same stick with us, and the tech career of your dreams will be

80
00:09:59,000 --> 00:10:05,000
right within reach. it's a pretty compelling promise across the Us.

81
00:10:05,000 --> 00:10:15,000
And Canada alone. The coding boot camp industry trained almost 25,000 people a year in 2,020, pulling in almost 350 million dollars.

82
00:10:15,000 --> 00:10:23,000
One of the claims made by coding boot camps is that they offer a novel solution for addressing some of the major problems within tech education.

83
00:10:23,000 --> 00:10:30,000
First, they're a lot shorter than a 4 year college degree. Second, they're supposed to have a much lower sticker. price.

84
00:10:30,000 --> 00:10:36,000
Third, they're supposed to be more accessible to groups that are traditionally excluded from the tech industry.

85
00:10:36,000 --> 00:10:41,000
And finally they're more agile than a university which is supposed to make them responsive to the needs of industry.

86
00:10:41,000 --> 00:10:52,000
With all of these factors combined, they're meant to be a new and ideal way to respond to the challenges of what some called the Fourth Industrial Revolution.

87
00:10:52,000 --> 00:10:57,000
But is any text scholar will tell you the likelihood of something tech related from the current moment.

88
00:10:57,000 --> 00:11:12,000
Being completely brand new is pretty unlikely. So I look to find the historical roots of the coding boot camp, and what I was able to find was surprising even to me, as I was reading through some work in the history

89
00:11:12,000 --> 00:11:18,000
of computing. I came across nathan Encemger's fantastic book about the software industry in the mid twentieth centuries.

90
00:11:18,000 --> 00:11:31,000
In it. He wrote about the prevalence of data. sorry electronic data processing schools for Edp schools in the 1,900 sixtys and seventys. Edp schools were vocational schools that offered short-term

91
00:11:31,000 --> 00:11:35,000
courses in computer programming and much like coding boot camps.

92
00:11:35,000 --> 00:11:39,000
These privately run schools, aim to train people for jobs in industry.

93
00:11:39,000 --> 00:11:49,000
After reading about Edp schools, I started looking for some primary historical materials about them, based on the sources available to me at the University, where I did my PHD.

94
00:11:49,000 --> 00:12:00,000
I was able to find a few newspaper articles. I also found ads for one franchise of schools in particular, which was the electronic Computer Programming Institute or the Ecpi.

95
00:12:00,000 --> 00:12:09,000
In the ads especially, I began to see some similarities between the sales, pitches of Edp schools and the sales, pitches of coding food camps, especially around.

96
00:12:09,000 --> 00:12:15,000
How easy programming can be, and the number of available jobs in the computing industry.

97
00:12:15,000 --> 00:12:28,000
I was hoping to find some more materials about Edp schools and traditional archives like the Charles Babbage Institute at the University of Minnesota, but because most Edp schools were either regional short-lived or

98
00:12:28,000 --> 00:12:34,000
both finding a's for the archives suggested that there wasn't a lot of archival material to be found.

99
00:12:34,000 --> 00:12:45,000
But then, thanks to the junk Mail collection of a computing pioneer named Ted Nelson, and the wonders of digitization, I stumbled upon something that I thought i'd never find which was an

100
00:12:45,000 --> 00:12:50,000
original recruitment brochure from the Ecpi, from the late 1,900 sixtys.

101
00:12:50,000 --> 00:13:03,000
The Ecpi booklet was only 16 pages long, but it was a crucial piece of evidence in my historical research that allowed me to make important links between the past and present of computing. education.

102
00:13:03,000 --> 00:13:12,000
It was really remarkable how many similarities there were between how Edp schools encoding boot camps framed themselves, and what they have to offer.

103
00:13:12,000 --> 00:13:24,000
First. both kinds of organizations framed a career in computing as accessible to anyone and a good way to go against the threat of automation.

104
00:13:24,000 --> 00:13:33,000
Edp schools also focused on the wide availability of computing jobs much like contemporary displayions of the Skills Gap.

105
00:13:33,000 --> 00:13:38,000
They also focused on computing as an inclusive career pathway for women and people of color.

106
00:13:38,000 --> 00:13:44,000
This was actually pretty remarkable for the 1,900 sixtys at a 10 students showcased for successful placements.

107
00:13:44,000 --> 00:13:49,000
Over half of them were women or people from minoritized ethnic groups.

108
00:13:49,000 --> 00:13:57,000
Finally, the Ecpi booklet underscored their links with industry Much like coding through camps, do today.

109
00:13:57,000 --> 00:14:03,000
Of course, the image that an organization presents to the world and its reality can be starkly different.

110
00:14:03,000 --> 00:14:11,000
So I then began to look into whether the image that Ecpi presented connected with other parts of the historical record.

111
00:14:11,000 --> 00:14:17,000
To do this, I turn to another fantastic Internet archive resource which was their digitized collection of computer world.

112
00:14:17,000 --> 00:14:27,000
To have this as a searchable digital archive was really incredible because it didn't contained some remarkably rich material that I probably wouldn't have found otherwise.

113
00:14:27,000 --> 00:14:39,000
One of the key directions that the Computer World Archive encouraged me to pursue in my research was the disconnect between the claims of inclusivity made about the computing industry at the time and the reality of the

114
00:14:39,000 --> 00:14:47,000
computing labor market. There was a ton of coverage in computer world about Edp schools in the 1,900 sixtys and seventys, and a lot of it highlighted.

115
00:14:47,000 --> 00:14:59,000
Many of the issues. With these organizations a series of articles published from 1,969 to 1,970, showed how these claims of accessibility were not necessarily so true.

116
00:14:59,000 --> 00:15:01,000
Despite a purported programmer storage.

117
00:15:01,000 --> 00:15:07,000
These articles illustrated how black programming trainees found it next to impossible to get hired for programming jobs.

118
00:15:07,000 --> 00:15:20,000
This was a very different story than the one told by the Ecpi booklet with the computer World Archive, also highlighted, was how problematic. So many of these schools were in Nathan encemger's.

119
00:15:20,000 --> 00:15:22,000
Book he discussed how many companies had employed A. No.

120
00:15:22,000 --> 00:15:27,000
Edp School graduate policy due to the variable quality of these schools.

121
00:15:27,000 --> 00:15:38,000
Articles that I found in computer world get clear insight into how these organizations operated, and the dire situations that some of their students were left in.

122
00:15:38,000 --> 00:15:43,000
So you may be asking yourself, Okay. So what does the history of Edp schools in the 1960 S.

123
00:15:43,000 --> 00:15:54,000
And seventies have to do with today. there's A reason that this history is worth paying attention to, and that's because it threatens to repeat itself in the current moment.

124
00:15:54,000 --> 00:16:04,000
The Ecpi booklet showed us that there are a lot of similarities between how coding boot camps and Edp schools presented both themselves and the potential benefits of learning to code.

125
00:16:04,000 --> 00:16:08,000
But the Computer World Archive shows us that there may be some other similarities as well.

126
00:16:08,000 --> 00:16:15,000
The story of Edp schools didn't end well either for many of the students, or for the schools themselves.

127
00:16:15,000 --> 00:16:18,000
They were shut down and liquidated by the mid 1,900 seventys.

128
00:16:18,000 --> 00:16:22,000
After a Federal investigation, cracked down on fraudulent vocational schools.

129
00:16:22,000 --> 00:16:27,000
Others just went out of business because no one wanted to go anymore.

130
00:16:27,000 --> 00:16:38,000
Of course, the future of coding boot camps has yet to be written, but there are some warning signs that there might be more similarities between food camps and Edp schools than we might hope one of my goals for this

131
00:16:38,000 --> 00:16:45,000
project was to point out how the past lives on in the present, so that the mistakes of the past can hopefully be avoided.

132
00:16:45,000 --> 00:16:56,000
If you've enjoyed this talk, I have a journal article coming out this fall in information, a culture that talks about the commonalities between coding boot camps, and Edp schools at length so please do keep an eye

133
00:16:56,000 --> 00:17:02,000
up. Thanks so much for your attention, and If You'd. like to get in touch, or put questions in the Q. A.

134
00:17:02,000 --> 00:17:11,000
Please do. Thanks so much, Kate. really appreciate that excellent overview. We have some questions that have come in

135
00:17:11,000 --> 00:17:16,000
I'm. gathering those if others in the audience have questions. Please do use the Q. and A.

136
00:17:16,000 --> 00:17:21,000
Feature to to drop those off and what we'll do now is move on to the next talk.

137
00:17:21,000 --> 00:17:26,000
And so up. Next is Tom Galley, from the University of Tokyo.

138
00:17:26,000 --> 00:17:31,000
Now Tom is is in Japan, and he will be watching this time shifted.

139
00:17:31,000 --> 00:17:34,000
So this is a wave to Tom. In the future he will be watching.

140
00:17:34,000 --> 00:17:40,000
This is in the past. i'm getting a little confused but let's let time explain.

141
00:17:40,000 --> 00:17:43,000
In this video. some of his research about Japan.

142
00:17:43,000 --> 00:17:47,000
So, Kaitlyn, you wanna you wanna take it away.

143
00:17:47,000 --> 00:17:54,000
Japan is now a modern country, not too different from many other places.

144
00:17:54,000 --> 00:18:03,000
But when it opened up to the world in the middle of the nineteenth century, Japan seemed exotic and mysterious to the first Western visitors.

145
00:18:03,000 --> 00:18:13,000
People who couldn't visit were even more curious about the country so to satisfy that curiosity, and to make some money.

146
00:18:13,000 --> 00:18:17,000
Those visitors wrote books about Japan for the people back home.

147
00:18:17,000 --> 00:18:22,000
Japan as they saw it, is a collection of excerpts from those books.

148
00:18:22,000 --> 00:18:42,000
They show how those visitors, mostly American, or British described the country and its people, the cities, the countryside, the clothing, the religions, the theater and festivals, even the smells they tell about writing in rick

149
00:18:42,000 --> 00:18:56,000
Shaw's attending weddings meeting prostitutes, experiencing earthquakes, Got the idea to create Japan as they saw it after I came across some of those old books at the Internet Archive.

150
00:18:56,000 --> 00:19:02,000
You see, I myself moved to Japan in 1,983 from California.

151
00:19:02,000 --> 00:19:17,000
I've lived here ever since, so it was fascinating for me to think about how my own first impressions of the country, you know, and what I had told my friends and family back home about it, compared with what people had

152
00:19:17,000 --> 00:19:27,000
written about Japan a century earlier, thought a collection of those earlier impressions might be interesting for others to read, too.

153
00:19:27,000 --> 00:19:42,000
So I first prepared a list of over 240 books about Japan at the Internet Archive, published between 1,855, and 1,912 went through those books and chose passages that seemed interesting.

154
00:19:42,000 --> 00:19:49,000
Or amusing, and I put them all on This website it's also available as an ebook.

155
00:19:49,000 --> 00:20:01,000
Each excerpt is linked to the original book at the Internet Archive, and, like those books at the Archive, the Japan, as they saw it, is free for anyone in the world to read.

156
00:20:01,000 --> 00:20:10,000
Thank you Once again, Internet Archive, from making this possible

157
00:20:10,000 --> 00:20:15,000
And I think so. You, Tom, for putting that video together, and if anyone has a question.

158
00:20:15,000 --> 00:20:22,000
Yes, We've already reached out to Tom and asked him. if we would create other videos for other parts of our collection, because I thought that was really great.

159
00:20:22,000 --> 00:20:30,000
Also stay tuned. Tom has A is gonna close out our show with another video on forgotten novels of the nineteenth century, which is really fun.

160
00:20:30,000 --> 00:20:43,000
Research. but up next we have a one of my favorite people on the planet a colleague who I've worked with for years in in a variety of ways, and i'm really pleased to welcome him to the to the

161
00:20:43,000 --> 00:21:02,000
stage the virtually to tell us about his research so that's a rod page from the University of Glasgow, talking about the bibliography of life over to you Rod

162
00:21:02,000 --> 00:21:06,000
Okay, Thank you very much for that lovely introduction. Chris.

163
00:21:06,000 --> 00:21:11,000
So i'm going to talk about the bibliography of life.

164
00:21:11,000 --> 00:21:15,000
So I guess what I should do is define what I mean by that.

165
00:21:15,000 --> 00:21:19,000
So of people look in the field of biodiversity and taxonomy.

166
00:21:19,000 --> 00:21:29,000
Have the stream of the bibliography of life and Basically, it's access to every taxonomic paper published on every species ever described.

167
00:21:29,000 --> 00:21:34,000
So to give you a sense of the scale. we Think there are probably about 2 million species described on the planet today.

168
00:21:34,000 --> 00:21:39,000
There's probably about 10 million in total so lots to be discovered.

169
00:21:39,000 --> 00:21:50,000
So we're thinking in terms of hundreds of thousands perhaps even a 1 million. also publications that describe the species times when we talk about this notion of a biography of life.

170
00:21:50,000 --> 00:21:55,000
Some people get upset because it seems to focus on on taxonomy.

171
00:21:55,000 --> 00:22:07,000
And what's so special that taxonomy what about all these major biomedical databases, such as pub, mid, for example, lots of information on on medicine, and so on to make the case that there's something special

172
00:22:07,000 --> 00:22:13,000
about about taxonomy, taxonomy and biodiversity in many ways.

173
00:22:13,000 --> 00:22:18,000
It's not big data. it's long data we have lots not so long.

174
00:22:18,000 --> 00:22:23,000
Tails. What you can see in the slide here on the right is

175
00:22:23,000 --> 00:22:29,000
Do a summary of the size of pages in Wikipedia for different mammal species.

176
00:22:29,000 --> 00:22:35,000
So some mammals, for example, lions and mice, charismatic animals, or medically important animals?

177
00:22:35,000 --> 00:22:39,000
Have we really large Wikipedia entries down the bottom of this chart?

178
00:22:39,000 --> 00:22:43,000
You can see the rest, literally thousands of pages of Wikipedia.

179
00:22:43,000 --> 00:22:55,000
On these mammals that are very, very small. So for many species on the planet, almost all that we know about in terms of the ecology, the morphology, what they do, where they are will come from the taxonomic

180
00:22:55,000 --> 00:23:03,000
literature, no taxing itself. she's also very much a subject for her.

181
00:23:03,000 --> 00:23:12,000
So we have some very large prominent journals, such as you Taxa, that published tens of thousands of new species descriptions.

182
00:23:12,000 --> 00:23:17,000
Also there is a very, very long tail in many cases of often very small journals.

183
00:23:17,000 --> 00:23:22,000
That are often very niche in terms of the textbook focus or the geographic focus.

184
00:23:22,000 --> 00:23:27,000
And again there will be a reservoir of lots of information about these species.

185
00:23:27,000 --> 00:23:33,000
So again, if we're going to get a nice comprehensive biography of life, we're gonna have to go hunting for those.

186
00:23:33,000 --> 00:23:39,000
Now those of you I guess, like like myself, from Chris, have been around for a while in this area.

187
00:23:39,000 --> 00:23:42,000
You might be thinking, hang on a second. This sounds already familiar.

188
00:23:42,000 --> 00:23:47,000
What about the biodiversity heritage library to beh shell isn't this what they're doing?

189
00:23:47,000 --> 00:23:52,000
And we had a really interesting presentation a couple of weeks ago, about them.

190
00:23:52,000 --> 00:24:02,000
Well, there are some overlap, but bh shell suffers from what I'm gonna call the Mickey Mouse Gap, which is this thing that the impact of copyright had a huge sort of dampening effect on vhs

191
00:24:02,000 --> 00:24:07,000
coverage. What i'm trying to capture in this diagram here is in blue.

192
00:24:07,000 --> 00:24:11,000
You can see for each decade how many publications are out there describing these species of animals.

193
00:24:11,000 --> 00:24:17,000
This is sort of lower bound on that, and in red is at the articles in Dhl.

194
00:24:17,000 --> 00:24:26,000
And you can see that the coverage will be actually dpped dramatically after about 1,923 when copyright in America, Hick kicks in and it's also starts to increase.

195
00:24:26,000 --> 00:24:30,000
A bit. But this is big area of blue and that's the stuff that I'm.

196
00:24:30,000 --> 00:24:37,000
After all these descriptions of species, mostly in the twentieth century, the out in Bhl.

197
00:24:37,000 --> 00:24:42,000
So once solution is to try and capture these these articles and put them some way safe.

198
00:24:42,000 --> 00:24:48,000
Now, a lot of this material has actually been digitized and It's really available in various websites.

199
00:24:48,000 --> 00:24:55,000
So I spent some time trying to grab this together. and i've sort of created almost a mini dhl on the Internet archive clickings, publications.

200
00:24:55,000 --> 00:25:05,000
They're full of beautiful photographs and species what's the information on geography and maps and climate, and also the people who study this species.

201
00:25:05,000 --> 00:25:10,000
It's just you try and gather as many of these publications as possible.

202
00:25:10,000 --> 00:25:16,000
I briefly talked about the the long tail of taxonomic publications, or these small journals.

203
00:25:16,000 --> 00:25:20,000
Many of these journals are as endangered as species, or they taxonomists themselves.

204
00:25:20,000 --> 00:25:24,000
There are generals that just simply vanish like this one of the left to the recent Italian journal.

205
00:25:24,000 --> 00:25:34,000
This is gone. they're also general set vanish but sort of come back as zombies taken over play by bad people on the right.

206
00:25:34,000 --> 00:25:44,000
There's a journal that used to publish on butterfly taxonomy the person who ran that eventually sort of came up was too hard to sort of maintain a taxonomic journal.

207
00:25:44,000 --> 00:25:48,000
Somebody else came along, took that domain name, and is now publishing things that anything.

208
00:25:48,000 --> 00:25:53,000
But butterfly taxonomy. and this is where the the way back machine is being incredibly spoiled.

209
00:25:53,000 --> 00:26:03,000
To try and retrieve. These are journals and the Content, and also discover history of some of these journals, which have been hijacked.

210
00:26:03,000 --> 00:26:11,000
One of those exercises just getting the content. one thing that that in a sense an archive, probably isn't particularly greater is is metadata.

211
00:26:11,000 --> 00:26:14,000
You can put quite a bit in. but for really detailed metadata.

212
00:26:14,000 --> 00:26:19,000
I've turned to wiki data so so what you Dixa has this extraordinary resource.

213
00:26:19,000 --> 00:26:30,000
It's a interesting interface and there's an enormous community of people contributing, So, for example, if you have an article like this is one that's in English and Chinese you can have titles in both

214
00:26:30,000 --> 00:26:36,000
languages. I think that really sort of motivates my use of Wiki data is to try and connect all these things together.

215
00:26:36,000 --> 00:26:43,000
So guess what i'm sort of aiming for ultimately is something a bit like this. there's there's a link.

216
00:26:43,000 --> 00:26:46,000
I think this is also gonna be in the the show notes.

217
00:26:46,000 --> 00:26:54,000
This is a little app that I built, that essentially talks to Wikipedia, and you can look at information, say, on species.

218
00:26:54,000 --> 00:27:03,000
On the left. you can get information on journals that publish papers on this species, and you can also then bounce to the the taxonomers, the people actually doing the research.

219
00:27:03,000 --> 00:27:12,000
So the goal is to have this kind of network of linked things, images of species, general articles, and the textbooks themselves, and these little thumbnails.

220
00:27:12,000 --> 00:27:15,000
You can see in these pictures here. they were coming from the Internet Archive.

221
00:27:15,000 --> 00:27:22,000
So these are articles that are freely available that people can read and see pictures of the species.

222
00:27:22,000 --> 00:27:33,000
So. So we're next for for this project I guess i'm very simple minded in a way one of it is just to try and get more content and get as much that content digitally preserved as possible which in an

223
00:27:33,000 --> 00:27:45,000
archive is one obviously to go down that and to use the way back machine to go and try and retrieve some of these journals that have come and gone, and the second stage is just to try and link with these things together.

224
00:27:45,000 --> 00:27:51,000
So we don't just have these digital artifacts that are sitting there preserved in some way. we can actually dance between them.

225
00:27:51,000 --> 00:28:05,000
Explore those connections. Thanks for much for your time. Thank you so much, Rod, and if you would stay on screen, and if Kate would come back on, I would love to do a quick round of questions I do have a

226
00:28:05,000 --> 00:28:12,000
question for each of you, and if again, if there are additional questions, do please drop them in the in the Q. A.

227
00:28:12,000 --> 00:28:17,000
But for Kate, the question is how much stuff is in Ted Nelson's junk mail collection.

228
00:28:17,000 --> 00:28:23,000
And how did you search through? It Seems like finding a needle in a haystack.

229
00:28:23,000 --> 00:28:28,000
So actually, I have to. I have to thank Google for that one.

230
00:28:28,000 --> 00:28:34,000
I was just looking for for information on Eddp schools.

231
00:28:34,000 --> 00:28:37,000
Actually I kind of don't was it I think it was just Google.

232
00:28:37,000 --> 00:28:47,000
I was looking for Ecpi. I was trying to find information on Ecpi, and I because you know the guide had been Ocr

233
00:28:47,000 --> 00:28:55,000
It showed up, and I just remember being like who's what is this junk mile junk mail collection like what is even happening.

234
00:28:55,000 --> 00:28:59,000
And then I clicked through, and I was like, oh, like this is miracle!

235
00:28:59,000 --> 00:29:06,000
Holy cow I can't believe that there's this call thing, and you know more than the actual booklet itself.

236
00:29:06,000 --> 00:29:14,000
There were other materials, including some news articles that had come with the packet that the sort of had encouraged sort of said, like, Oh, you know.

237
00:29:14,000 --> 00:29:20,000
Look how great, you know, learning to code or you know computer learning computer programming is.

238
00:29:20,000 --> 00:29:25,000
And you're gonna make so much money and all of that so

239
00:29:25,000 --> 00:29:29,000
I have not actually seen enough of, I think the junk mail file but

240
00:29:29,000 --> 00:29:37,000
I should definitely just blocking one day that's I I bet many of the researchers here in the audience have the same experience. I know.

241
00:29:37,000 --> 00:29:50,000
I did, you know, doing a Google search for something kind of upcoming in your in your discipline, and like, lo and behold, there it is somewhere at the Internet Archive, like I wouldn't have never thought to go look in the ted nelson junk

242
00:29:50,000 --> 00:29:53,000
mail collection for this thing that I needed. Yeah, thank you for that, Kate.

243
00:29:53,000 --> 00:29:59,000
A question for you rod from we'll we'll pick this one up here from from martin cafatovic.

244
00:29:59,000 --> 00:30:12,000
Now I know that you're not a a copyright lawyer. but I bet you have some opinions on what the question is that Martin is asking, which is that current global copyright Laws are terrible for science and for culture, But people go to

245
00:30:12,000 --> 00:30:16,000
jail, or worse for violating them. Do you have any thoughts on how institutions can risk?

246
00:30:16,000 --> 00:30:31,000
Manage that. Well, that's that's that's a question I guess what one thing did to point out so most of the content that i'm going after I mean i've thought about is it's quote unquote freely

247
00:30:31,000 --> 00:30:34,000
available, or some of that is explicitly, freely licensed.

248
00:30:34,000 --> 00:30:44,000
So I think I know there's an element. of If you model is negotiate with the providers to sign agreements to get the content which I guess is the way Bh ol does which is all proper that

249
00:30:44,000 --> 00:30:54,000
mitigates the risk. i've sort, of taken the approach well, if it's not kind of like behind the pay walls or not expressly saying, do not have this then, we can at least try and make copies of it so.

250
00:30:54,000 --> 00:30:58,000
That it doesn't go away and again we've seen repeatedly. Yeah.

251
00:30:58,000 --> 00:31:01,000
By the time we could have had these negotiations.

252
00:31:01,000 --> 00:31:06,000
Some of these journals are just gone. so I I take Martin's point.

253
00:31:06,000 --> 00:31:14,000
I guess that's an interesting question. for the interactive itself it's an institution that presumably tries to mitigate the risk of dealing with this issue.

254
00:31:14,000 --> 00:31:24,000
And leila bailey from the policy expert at the Internet Archive has entered the chat and and dropped off some some thoughts as did Brewster.

255
00:31:24,000 --> 00:31:27,000
So we'll let that conversation continue on asynchronously.

256
00:31:27,000 --> 00:31:31,000
Kate Rod, Thank you much for your presentations today.

257
00:31:31,000 --> 00:31:35,000
If you have additional questions for either one of them drop them into the into the Q.

258
00:31:35,000 --> 00:31:45,000
And a and we'll pick up with our next Speaker and so I would like to welcome Michelle Alexopolis to the to the to the screen today.

259
00:31:45,000 --> 00:31:48,000
And Michelle is gonna tell you about her research.

260
00:31:48,000 --> 00:32:02,000
Looking at expressions on the Federal chairs about the members of the Board of Federal Reserve Board chairs, and how that their expressions affect financial markets.

261
00:32:02,000 --> 00:32:07,000
So over to you, Michelle. Thank you very much.

262
00:32:07,000 --> 00:32:14,000
I would like to thank the organizers, obviously, for putting together such an interesting eclectic group of papers.

263
00:32:14,000 --> 00:32:19,000
So I would. as Krista said, I wanna talk about more than words.

264
00:32:19,000 --> 00:32:32,000
Basically, this is an analysis of the Federal reserve board Chairs communication during Congressional testimony, based on a project that I'm doing with members of Bank of Canada specifically Asian Fen Han Lesky pressed off and Shu Jang the link

265
00:32:32,000 --> 00:32:38,000
to our newly released working paper is available there now in terms of the product data sources.

266
00:32:38,000 --> 00:32:42,000
This wouldn't have happened unless we had access to a tremendous amount of resources.

267
00:32:42,000 --> 00:32:47,000
We're looking for audio video as well as textual from the Internet Archive.

268
00:32:47,000 --> 00:32:58,000
We've primarily used the Tv archive content these are videos from Cspan, Cnbc Bloomberg Tv and the house and Senate videos and where there is polls in that coverage.

269
00:32:58,000 --> 00:33:07,000
We've then turned to c-span archives Youtube videos, textual sources that would give us transcript information.

270
00:33:07,000 --> 00:33:16,000
And what we did is we analyzed all of this and then tried to blend it with basically tick-by tick stock data that came from places like Warden and refuitive etc.

271
00:33:16,000 --> 00:33:19,000
So the over overview of the project is pretty straightforward.

272
00:33:19,000 --> 00:33:23,000
What we wanted to do is examine the importance and the impact of the Us.

273
00:33:23,000 --> 00:33:35,000
Federal reserve communication, we're going to measure aspects of the communication using Ai natural language processing and other tools, and that we want to test if and how the words chosen their body language and their tone of voice is

274
00:33:35,000 --> 00:33:42,000
actually affecting markets. So when we think about policy communication, I understand we have a very large audience.

275
00:33:42,000 --> 00:33:51,000
When I talk about monetary policy and policy communications, i'm talking about things such as interest rate movements for the policy rate for guidance, quantitative easing or bank balance.

276
00:33:51,000 --> 00:33:57,000
Sheet operations. These things have very large impacts on the welfare of businesses and individuals.

277
00:33:57,000 --> 00:34:02,000
In our economy, not just in the United States, but because they're so large around the world.

278
00:34:02,000 --> 00:34:06,000
Now these policies are often fairly complex now to be effective.

279
00:34:06,000 --> 00:34:09,000
Central bank communications need to be accurate. They need to be clear.

280
00:34:09,000 --> 00:34:14,000
They need to be perceived as credible. and they need to reach their target audience.

281
00:34:14,000 --> 00:34:22,000
And it's true that when the central bank takes actions or releases communications, we at all are not necessarily watching the actual release.

282
00:34:22,000 --> 00:34:35,000
But we're also getting our information through media. so we have 2 different types of channels where we get this information, it could be direct communication, such as the press conferences themselves or through watching things like the testimony or even the

283
00:34:35,000 --> 00:34:47,000
press releases, such as the Fafc statements that were released last week, or we can have the indirect types of communications that we often see captured by the Internet archives Tv archive which deals with the coverage on

284
00:34:47,000 --> 00:34:53,000
Cnbc fox news, or even things like Bloomberg Tv.

285
00:34:53,000 --> 00:34:57,000
This is also then amplified by sort of traditional things, such as the New York Times and and Twitter.

286
00:34:57,000 --> 00:35:06,000
Now, when we think about communication communication, is obviously more than words, and this has been popularized for a very long time ever since the 1,900 seventys, I'm.

287
00:35:06,000 --> 00:35:13,000
Sure some of you have heard that the statement that over 90% of communication it comes from your body, language, and tone of voice.

288
00:35:13,000 --> 00:35:22,000
Now, although the Morabian studies are probably not getting exactly the the right numbers, in terms of which quantities the idea is pretty much accepted.

289
00:35:22,000 --> 00:35:32,000
So you can see here even a few pictures of Ben Bernanke and Janet Yellen, who are 2 of the Federal reserve chairs, and you can see sometimes they look very relaxed, and sometimes they look a lot more

290
00:35:32,000 --> 00:35:37,000
perplexed or concerned. So we want to think about how that actually influences people.

291
00:35:37,000 --> 00:35:44,000
And why does body language potentially matter body, language, and tone of voice, could matter through a direct channel?

292
00:35:44,000 --> 00:35:50,000
So a traitor or us individuals may be watching directly and then we take actions based on what it is we're hearing.

293
00:35:50,000 --> 00:35:57,000
We're either actively looking for something or it may affect how we feel about something or sentiment, or confidence.

294
00:35:57,000 --> 00:36:08,000
If we're, looking at indirect channels because these get picked up and amplified through other Tv stations and other news outlets, what we can basically get is that the fed has a communication the journalists.

295
00:36:08,000 --> 00:36:11,000
Or the analysts are watching it, and then they're releasing information.

296
00:36:11,000 --> 00:36:15,000
And now those of us who are receiving that information, it could be broken.

297
00:36:15,000 --> 00:36:19,000
Telephone may not be, but we'll take actions upon that themselves.

298
00:36:19,000 --> 00:36:29,000
So you can see here, for example, a quote that came from the Cnbc Street signs that was talking about when Janet Yellen looked insecure when she looked frustrated when she looked angry and even when Ben

299
00:36:29,000 --> 00:36:34,000
bernanke's voice potentially got shaky at at various times.

300
00:36:34,000 --> 00:36:43,000
And then there's also the impact of algorithmic traders, where now we have an automated channel where they may be taking a lot of these signals, analyzing them and trading on that behalf.

301
00:36:43,000 --> 00:36:45,000
So what our research tries to do is look at these different channels.

302
00:36:45,000 --> 00:36:48,000
We're not going to take a stand on which one is most important.

303
00:36:48,000 --> 00:36:58,000
But we want to take a look at fed share testimonies, and this is a little different than some of the other papers that have been looking at fed communications because we're focusing on these 32 testimonies which will

304
00:36:58,000 --> 00:37:03,000
expand over the summer and we're looking at the 3 different dimensions of communications.

305
00:37:03,000 --> 00:37:14,000
All at once. Now, what's interesting about these communications is There's also a prepared portion of this, and we have unscripted question and answers and people can obviously respond to those things in different ways.

306
00:37:14,000 --> 00:37:19,000
We will also have the responses from the Senators and the Congressional Representatives.

307
00:37:19,000 --> 00:37:27,000
Now why testimonies you might ask well first of all they're widely washed by fed watchers and investors.

308
00:37:27,000 --> 00:37:31,000
It's covered by a lot of news media it doesn't occur on the same day as an actual policy announcement.

309
00:37:31,000 --> 00:37:35,000
So you can focus a little bit more on disentangling just the communications component.

310
00:37:35,000 --> 00:37:44,000
And, of course, a high quality Tv footage and transcripts that are available from the archives has made this actually a doable thing at the moment.

311
00:37:44,000 --> 00:37:46,000
So the structure of the semi-anual testimony.

312
00:37:46,000 --> 00:37:48,000
These things are about 2 to 3 h, and they come in pairs.

313
00:37:48,000 --> 00:37:54,000
So the semi-annual testimonies you end up with one before the House and one before the Senate.

314
00:37:54,000 --> 00:38:00,000
Banking committee. You have the prepared remarks released before things be in. Then you'll have some welcoming statements.

315
00:38:00,000 --> 00:38:05,000
The prepared remarks in the Q. and A. What we can do then is we can disentangle a different components.

316
00:38:05,000 --> 00:38:09,000
Here we have transcripts which are going to give us the actual words that are spoken.

317
00:38:09,000 --> 00:38:16,000
We have the audio component, and then we have the video component which will allow us to look at how facial expressions are formed.

318
00:38:16,000 --> 00:38:21,000
So here the 3 different ways, basically from the text without getting into details.

319
00:38:21,000 --> 00:38:31,000
We're using a fine-trained bert model which is a language model that sort of stayed in the art right now from the voice we're basically forcing alignments with transcripts and we're

320
00:38:31,000 --> 00:38:34,000
extracting pitch for each speaker. with a tool called Pr.

321
00:38:34,000 --> 00:38:41,000
And for the facial expressions we're using things that will be used in laboratories in psychology.

322
00:38:41,000 --> 00:38:54,000
So, faith, reader, for example, we're using microsoft azure, and we're matching these to look at facial action units and then using mappings that have been done with psychologists before into figuring out what kind of

323
00:38:54,000 --> 00:38:57,000
negative emotions might be being expressed by the Federal reserve chairs.

324
00:38:57,000 --> 00:39:00,000
During this time. Now all in all, we take all of this.

325
00:39:00,000 --> 00:39:04,000
We then have to mash it in with financial market. data might be very careful about the timing of this.

326
00:39:04,000 --> 00:39:10,000
So we use the Cnbc. live coverage and We're watching what the values of the S.

327
00:39:10,000 --> 00:39:14,000
And P. 500 are in order to do some of that crosswalk in the timestamps.

328
00:39:14,000 --> 00:39:28,000
Then what we do is we basically create this giant database We're going to use local projection regressions and difference in different type projections in order to look at the impact of these different channels of emotions on the outcomes which would

329
00:39:28,000 --> 00:39:32,000
be, say, the change in the S. and p 500 or what's happening with the vix.

330
00:39:32,000 --> 00:39:37,000
Now, what we typically find in this is just one of the examples is, yes, indeed, the S.

331
00:39:37,000 --> 00:39:40,000
And P. 500, and the vix tend to respond in the ways that you might expect.

332
00:39:40,000 --> 00:39:53,000
So if they're actually relaxed and happy not negative, you end up seeing the values of the stock markets going up in the short run, and the value of the vix, which is usually considered to be a fear index, going down Now, what

333
00:39:53,000 --> 00:40:05,000
you also see, is there are differences across different topics, and, as you might expect, the ones that the fed has most oversight or control on such as monetary policy will have the largest impact on markets over the time period Now for some reason findings

334
00:40:05,000 --> 00:40:11,000
today. I did promise, Chris. I would try to get in on Time is what we did find is these different dimensions.

335
00:40:11,000 --> 00:40:18,000
They don't all act at the same time in a highly correlated fashion, text, voice, and facial emotions, though, do tend to move financial markets.

336
00:40:18,000 --> 00:40:24,000
I showed you a couple of examples for the short run responses. but the impact seems to grow stronger in the days following these testimonies.

337
00:40:24,000 --> 00:40:34,000
There's differences in the amount of soft information how they can differ across topics, and we also are starting to see some patterns that we're going to be verifying over the summer as to whether or not some of these

338
00:40:34,000 --> 00:40:37,000
responses differ from chat, fed, different fed shares.

339
00:40:37,000 --> 00:40:44,000
We also have the Congressional members emotions, and sometimes they will significantly move markets as well. alright.

340
00:40:44,000 --> 00:40:47,000
So with that, I say, thank you comments and questions are welcome again.

341
00:40:47,000 --> 00:40:51,000
There's a link to working paper and you can reach out to me via email.

342
00:40:51,000 --> 00:40:57,000
I'll be happy to have a conversation thank you very much, thank you, Michelle.

343
00:40:57,000 --> 00:41:00,000
We? We have gotten a couple of questions and come in to me.

344
00:41:00,000 --> 00:41:04,000
Again, if you have additional questions for michelle please drop them off in the Q. A.

345
00:41:04,000 --> 00:41:14,000
But what we'd like to do now just move on to our next speaker. and what again we'll we'll pick up questions with both salute and michelle in just a minute.

346
00:41:14,000 --> 00:41:19,000
But today, right now, I want to welcome. So wot along to the screen.

347
00:41:19,000 --> 00:41:30,000
So we works for the Internet archive he's part of the way back machine team doing amazing things he's gonna tell you walk you through some of the work that he's doing on metadata behind the scenes with

348
00:41:30,000 --> 00:41:36,000
our web archive. So So we would please over to you.

349
00:41:36,000 --> 00:41:42,000
Allow me to share my screen

350
00:41:42,000 --> 00:41:46,000
Take a while

351
00:41:46,000 --> 00:41:54,000
Okay, This top one share this is my screen. visible.

352
00:41:54,000 --> 00:42:07,000
Looks great, awesome, Thank you So i'm solid alums a webinar data scientist or webex machine at the Internet Archive, and to see on my screen is the help text of a tool called cdx

353
00:42:07,000 --> 00:42:12,000
summaries I mean my current working breath. We have a file for samples or Cds or Gc.

354
00:42:12,000 --> 00:42:18,000
Which is a compressed index of a bunch of web app type files.

355
00:42:18,000 --> 00:42:22,000
But we don't know what was in those that archives 5 for a while.

356
00:42:22,000 --> 00:42:28,000
Hopefully. when we run this index through our cdx summary tool, we'll get to know what is entirely there.

357
00:42:28,000 --> 00:42:32,000
So I'm going to run this command here and it'll take a while to complete.

358
00:42:32,000 --> 00:42:37,000
So we'll come back to it later. let's let's go to Internet Archives home.

359
00:42:37,000 --> 00:42:44,000
Page. Yeah, we see a bunch of collections listed. Some are collections of books and audio and video.

360
00:42:44,000 --> 00:42:55,000
Some are web collections let's open one of these collections, and in this case it is Ukrainian culture on heritage collection.

361
00:42:55,000 --> 00:43:00,000
It's called through this collection to see some interesting colorful thumbnails.

362
00:43:00,000 --> 00:43:07,000
The meaningful titles, and you have a rough idea of what to see what to get in in this collection.

363
00:43:07,000 --> 00:43:22,000
Now let's go to a web. collection which is a collection of work, file an arbitrary number of web captures are packaged together, and in this case we see a list of tombstones here as

364
00:43:22,000 --> 00:43:28,000
thumbnail and titles look very similar as if they were generated by machine.

365
00:43:28,000 --> 00:43:37,000
It doesn't tell a lot there is an about tab in this collection view. But we don't see much there are few lines of metadata here and at all.

366
00:43:37,000 --> 00:43:40,000
And this is the problem that you are going to start here.

367
00:43:40,000 --> 00:43:50,000
In the next few days or weeks, when the testing is done in, the changes are merged to the main site.

368
00:43:50,000 --> 00:43:59,000
This case will look something like that. Well, we'll get to know how many you capture are in this collection.

369
00:43:59,000 --> 00:44:05,000
How many unique uris were captured. how many of those youri's belong to say Html pages are 200.

370
00:44:05,000 --> 00:44:15,000
Okay, for example, or how many of those Url have 0 cats and 0 queries? that is, root pages versus deepling.

371
00:44:15,000 --> 00:44:19,000
We can also get simple and spread out the collection.

372
00:44:19,000 --> 00:44:23,000
So this collection started in 2,016 and stop there.

373
00:44:23,000 --> 00:44:26,000
No activity in 2,017, and then it started again.

374
00:44:26,000 --> 00:44:32,000
2,018, and since then it's been constant constantly archiving every single month.

375
00:44:32,000 --> 00:44:39,000
Then we also have a bunch of top folks that are contributed into this collection with maximum number of capture.

376
00:44:39,000 --> 00:44:44,000
And one cool thing that it the tool brings up is a bunch of sample.

377
00:44:44,000 --> 00:44:51,000
Uris from this collection. so wackf is a bundle of a number of your either.

378
00:44:51,000 --> 00:44:59,000
You don't know what is inside. in there and this allows us to kind of, you know, click on one of these links, and see how, when those pages are archived, so it helps us doing.

379
00:44:59,000 --> 00:45:06,000
Qa. for example, all these numbers up here. They also allow us to gain insight, and learn.

380
00:45:06,000 --> 00:45:12,000
Sometimes, if the numbers are really off, we look like there was some problem in our calling process and go back and pick them.

381
00:45:12,000 --> 00:45:16,000
So how this whole page was rendered is basically backed by A.

382
00:45:16,000 --> 00:45:25,000
Json file, and this Jason file is generated using a command line tool called Cdx Summary, and we have open source.

383
00:45:25,000 --> 00:45:28,000
This tool, so you can use this tool to generate human readable.

384
00:45:28,000 --> 00:45:40,000
You know, summaries or machine Read people, Json files, and if you have a Json file you can render it in Sdml using this web component that we made available.

385
00:45:40,000 --> 00:45:54,000
So with that, let's go back to the command line and See what happened to the local cdf file that we had, and it turned out it did generate a very nice human readable somebody off the the cdfs that we

386
00:45:54,000 --> 00:46:00,000
had. Now we get to know what is inside in there, who, when it was captured, and how many catches up there, and so on, and so forth.

387
00:46:00,000 --> 00:46:07,000
And we can even click on one of these links to to open them in a browser and see a hard for that type.

388
00:46:07,000 --> 00:46:15,000
But by demonstrating it on command line. What we really illustrated is this: pool is not tied to Internet archives.

389
00:46:15,000 --> 00:46:18,000
It is an independent tool. Actually, we are just using an internal archive.

390
00:46:18,000 --> 00:46:22,000
Anyone can use it in their own Webcastle collection.

391
00:46:22,000 --> 00:46:32,000
So with that I will recap. We basically created a command line tool and a companion, but the component to summarize the web collections.

392
00:46:32,000 --> 00:46:36,000
We release these tools under open source license.

393
00:46:36,000 --> 00:46:42,000
And we we use these tools to enrich our own web connections.

394
00:46:42,000 --> 00:46:47,000
Begin inside from generated staff to improve our callers.

395
00:46:47,000 --> 00:46:54,000
Finally, we demonstrated that these tools are not explosive to the Internet archives, so anyone can use them to their own.

396
00:46:54,000 --> 00:46:59,000
The archival collection, and learn more more about what they are having.

397
00:46:59,000 --> 00:47:05,000
With that I will see you in Qa. and thank you so much for with me.

398
00:47:05,000 --> 00:47:10,000
Thank you so much. so we would. You know that lots of comments and questions coming in in the chat.

399
00:47:10,000 --> 00:47:14,000
So I think we'll we'll have a good question for you in just a minute.

400
00:47:14,000 --> 00:47:25,000
But up next we have a art rhino and I had the chance of watching this video, and I think it's gonna blow your mind.

401
00:47:25,000 --> 00:47:29,000
It's a little, Willie wonka and just charming and wonderful.

402
00:47:29,000 --> 00:47:39,000
So arts gonna show you in this video kind of the scanning robot that he's built alongside one of the Internet archives.

403
00:47:39,000 --> 00:47:41,000
Scanning frames the tt scanner.

404
00:47:41,000 --> 00:47:47,000
So, Caitlin, let's roll the the video hi there!

405
00:47:47,000 --> 00:47:52,000
My name is our rhino. I work at the University of Windsor and Ontario, Canada, and I'm.

406
00:47:52,000 --> 00:47:57,000
On the board of a nonprofit organization called our Digital World, or Odw.

407
00:47:57,000 --> 00:48:02,000
My library acquired a very nice tabletop scanning machine from the Internet Archive.

408
00:48:02,000 --> 00:48:09,000
Years ago. This unit currently has good availability and i've become interested in scanning the back file for a major paper collection.

409
00:48:09,000 --> 00:48:13,000
A lot of the material is well positioned for scanning.

410
00:48:13,000 --> 00:48:16,000
Some of it can be disassembled, and the pages can be fed through.

411
00:48:16,000 --> 00:48:23,000
Our Xerox can feature. However, there's a lot of spit binding. I've been trying to find a way to scan it without compromising the binding.

412
00:48:23,000 --> 00:48:33,000
Here are the components i've used to pull together. on a Mag scanning most of which were already on hand from Odw's work on Diy Mic Films scanners.

413
00:48:33,000 --> 00:48:40,000
One new edition as an inflatable snow tube which came from the end of season table at a local hardware store.

414
00:48:40,000 --> 00:48:45,000
The year. Mattress pump inflates the snow tube, and a vacuum cleaner holds on to the page.

415
00:48:45,000 --> 00:48:51,000
The device in the front that looks like a car is the embod, or in Block Inbot.

416
00:48:51,000 --> 00:48:54,000
My embod is several years old, and I use the in-bloc.

417
00:48:54,000 --> 00:49:02,000
Id on the web for controlling it. There are wireless and bluetooth options for newer models. That would be more elegant for this kind of thing.

418
00:49:02,000 --> 00:49:07,000
However, this worked well enough, and it seems to do the trick for what we want to do.

419
00:49:07,000 --> 00:49:16,000
Probably the biggest drawback of all this is the noise from the air mattress pumping the vacuum cleaner, which you can't hear here, but otherwise i'm pretty happy with this approach.

420
00:49:16,000 --> 00:49:21,000
It takes about 30 to 45 min to scan the typically 100 pages in a title.

421
00:49:21,000 --> 00:49:29,000
You can find way more sophisticated and faster devices on, Youtube and it would be cool to make something more general purpose and robust.

422
00:49:29,000 --> 00:49:37,000
But this is one approach for automatically scanning material on the tabletop scribe that is hopefully useful for others.

423
00:49:37,000 --> 00:49:50,000
Thanks for listening. So I I have it on good authority that art does a lot of really interesting projects, and is, as you can see, as a little bit of a tinkerer.

424
00:49:50,000 --> 00:49:55,000
And just what creativity I think that someone mentioned like a rotisserie chicken.

425
00:49:55,000 --> 00:50:00,000
Turning thing next to a microfilm scanner, anyway.

426
00:50:00,000 --> 00:50:11,000
Interesting stuff there. I would like. to welcome or i'd like to bring Michelle, and so we back to the to the screen for a couple of questions.

427
00:50:11,000 --> 00:50:15,000
And so the first question that I have, if you wanna turn your cameras on.

428
00:50:15,000 --> 00:50:18,000
Thanks, Michelle. The The question that I have is for you.

429
00:50:18,000 --> 00:50:28,000
Are you data points, or related data that you'd like to including your research that simply weren't or aren't available in digital or computable form.

430
00:50:28,000 --> 00:50:33,000
And so they's a great question. there are holes in the collection.

431
00:50:33,000 --> 00:50:47,000
Unfortunately, the Internet archive hasn't gone back quite far enough for for what we'd love to do which is kind of go back sort of to the beginning to include a lot more of the data from say green spans time period or or even

432
00:50:47,000 --> 00:50:57,000
people passed it passed on that. and sometimes some of the audio files are some other forms of communications that could have happened around the time.

433
00:50:57,000 --> 00:51:09,000
So some of that's not available some we're still trying to dig through some archives to see whether or not they've been put someplace else or sort of miss misplaced on on some of that but there

434
00:51:09,000 --> 00:51:15,000
is the unfortunate thing that I mean, why the Internet archives so important that things have gotten destroyed over time.

435
00:51:15,000 --> 00:51:18,000
So we would really like to have it as inclusive as possible.

436
00:51:18,000 --> 00:51:27,000
We'd also like to look at some of the issues between Janet Yellen and other female fed authorities, and whether or not there's different responses on them.

437
00:51:27,000 --> 00:51:33,000
And just sometimes the the video feeds and everything like that just aren't there right now.

438
00:51:33,000 --> 00:51:42,000
So those are the primary ones. we're getting much better at getting stock data and everything coming in. So that's not quite so much of a problem fascinating.

439
00:51:42,000 --> 00:51:45,000
Thanks so much for that. So we would a question for you.

440
00:51:45,000 --> 00:51:49,000
The comment is, Thank you for making your cdx tool open source.

441
00:51:49,000 --> 00:51:55,000
What other projects would you envision using your tool? Oh, interesting!

442
00:51:55,000 --> 00:52:07,000
So One immediate application. I was thinking the other day is there is an emerging format called wacky,

443
00:52:07,000 --> 00:52:12,000
Which is like bundling a bunch of work files with their indices and all that stuff.

444
00:52:12,000 --> 00:52:21,000
I think this tool can integrate well inside in there so it's someone kind of you know, loads the faxie file in a browser or they can see what what really is inside.

445
00:52:21,000 --> 00:52:27,000
And it has been a challenge like I mean director of the payback machine.

446
00:52:27,000 --> 00:52:34,000
He often asked me, Okay, So here is this item we collected what what is inside in there, and you know, is the tool.

447
00:52:34,000 --> 00:52:40,000
Go run a game state, and you will have, like a bunch of few models that you can play with and explore from there again.

448
00:52:40,000 --> 00:52:47,000
I mean, this is both a a statistical insight as well as kind of new exploratory.

449
00:52:47,000 --> 00:52:53,000
Another option. i've been looking forward for is to kind of know

450
00:52:53,000 --> 00:52:58,000
These random url that we pull out Sundays, and those are not just random.

451
00:52:58,000 --> 00:53:12,000
They have some other criteria to that we can use those to generate thumbnails and have like a slide show, or something like that to to have a more visual you know understanding of what's been pied in a

452
00:53:12,000 --> 00:53:15,000
collection and use it as a as a way to kind of tell.

453
00:53:15,000 --> 00:53:21,000
Tell a story about a collection that's that thank you for that.

454
00:53:21,000 --> 00:53:28,000
For that explanation. I I understand from a from a message from Michael Nelson in chat that you're gonna be teaching a course next semester.

455
00:53:28,000 --> 00:53:38,000
Oh, yes, it will be webinar design. So yeah, my students will be competing with say, Engine X and Apache and some of other fancy books are work.

456
00:53:38,000 --> 00:53:46,000
Hope to have a more you know. Compliant with server with rice or standards.

457
00:53:46,000 --> 00:53:52,000
Then then some of these other Well, known what's that what's that.

458
00:53:52,000 --> 00:53:57,000
Say, Well, good luck to you in the in the course, and

459
00:53:57,000 --> 00:54:00,000
Your students are in excellent hands for sure, Michelle.

460
00:54:00,000 --> 00:54:07,000
So we would thank you so much for your presentations today, and we'd like to keep moving through the through today's talks.

461
00:54:07,000 --> 00:54:12,000
If you had additional questions for Michelle, or so we would please do drop them off into the Q.

462
00:54:12,000 --> 00:54:23,000
And a but what i'd like to do now is welcome Spencer tourine to the to the screen from Thompson Reuters special services and spencer's gonna talk about the work that he's done

463
00:54:23,000 --> 00:54:27,000
with automatic hashtag hierarchy generation and I can't wait to hear that.

464
00:54:27,000 --> 00:54:32,000
So spends her over to you

465
00:54:32,000 --> 00:54:38,000
Yep

466
00:54:38,000 --> 00:54:47,000
Okay, hello, welcome everyone. Yeah. So i'd like to talk about well, what Chris just said.

467
00:54:47,000 --> 00:54:54,000
So. The idea is, you know we have all these tags hashtags on Twitter, and the question is, do they have a hierarchy?

468
00:54:54,000 --> 00:55:02,000
Is there structure to them and i'll talk a little bit about the difference between ontologies and folksonomies really quickly.

469
00:55:02,000 --> 00:55:08,000
So we're all probably, if not familiar with ontologies by name, at least by concept.

470
00:55:08,000 --> 00:55:14,000
So we have things like the dictionary, shall we say in the dictionary there are very curated terms.

471
00:55:14,000 --> 00:55:20,000
There are rigorous definitions of these terms, and there are, you know, well-defined relations between these terms.

472
00:55:20,000 --> 00:55:26,000
So an example is, a foot has 5 toes, and as part of the human body there are many more definitions of what a foot is.

473
00:55:26,000 --> 00:55:30,000
You can see on the right, and I think on the right is just some of them.

474
00:55:30,000 --> 00:55:36,000
But we have to compare that with the difficulty of looking at let's say twitter in this respect.

475
00:55:36,000 --> 00:55:44,000
So on Twitter. it's been called a folks on me so these hashtags that are used are very arbitrary.

476
00:55:44,000 --> 00:55:50,000
Their definitions are circumstantial. there are very undefined relationships to them.

477
00:55:50,000 --> 00:55:58,000
I just created a a hashtag down there, 2 hashtags done there, and if I use them on Twitter, they would now exist in the record.

478
00:55:58,000 --> 00:56:03,000
What do some of those hashtags even mean I don't even know?

479
00:56:03,000 --> 00:56:07,000
Really. I just put up some nonsense just to illustrate the fact that I could do it.

480
00:56:07,000 --> 00:56:18,000
So yeah. if we want to put some structure to these hashtags, we have to be able to ask at least the minimal question, which is, you know, are there hashtags which are more general than others shall we

481
00:56:18,000 --> 00:56:22,000
say, is hashtag dog more general than Dalmatian.

482
00:56:22,000 --> 00:56:27,000
If dog in our, in our anthological thinking, is more general than dalation, Is that true?

483
00:56:27,000 --> 00:56:31,000
On Twitter is Ai has to take Ai more general than hashtag machine learning.

484
00:56:31,000 --> 00:56:45,000
What does General mean on Twitter? and ontology. would say that a a general word is perhaps due to the breadth of applicability of the term, and that, again would be by the definition.

485
00:56:45,000 --> 00:56:59,000
There's nothing to say that you could or should use a term very often, but by the definition of the term, you could say that the term is general with folksonomies, since there's no real clear definition for these

486
00:56:59,000 --> 00:57:06,000
things. We probably have to take the question in a different direction, and talk about the breath of application rather than the breath of applicability.

487
00:57:06,000 --> 00:57:10,000
So we're gonna ask certain contextual questions like you know when was he used?

488
00:57:10,000 --> 00:57:17,000
Where was he who used it? How is it used? So we look at context as proxies for generality?

489
00:57:17,000 --> 00:57:21,000
So there's a classic one which would be the popularity context.

490
00:57:21,000 --> 00:57:27,000
If a lot of people are using it that might indicate it's for general use, there are lots of temporal contexts.

491
00:57:27,000 --> 00:57:36,000
So here are some examples. hashtag disease would be you know, not related to a specific event, although Covid 19 is. It's a big deal.

492
00:57:36,000 --> 00:57:44,000
It's a big event. But it's none loss of an event, and in some years i'm probably will see that Hashtag die out.

493
00:57:44,000 --> 00:57:48,000
You can have a general holiday which could happen, you know, any time of the year.

494
00:57:48,000 --> 00:57:53,000
Halloween is generally one time a year you could have hashtag work which occurs throughout the week.

495
00:57:53,000 --> 00:58:03,000
Tgif might be a Friday thing, and you could certainly have the hashtag food any any time of the day and breakfast maybe more so in the morning.

496
00:58:03,000 --> 00:58:12,000
And then the all important semantic contexts. so you know, if you have a hashtag that co-occurs with many other different hashtags.

497
00:58:12,000 --> 00:58:15,000
This is sort of the state of the art of the moment. You know which other?

498
00:58:15,000 --> 00:58:19,000
How many other different, unique hashtags is a hashtag used with.

499
00:58:19,000 --> 00:58:25,000
So if if very many, then you could call it general, and if not very many, then maybe not so general.

500
00:58:25,000 --> 00:58:33,000
And then we can also look at topics which you might consider to be groups of hashtags is used within a group or outside the group, and also tokens and words.

501
00:58:33,000 --> 00:58:47,000
So is it used with many different ideas. or not and the way we measure all this is basically we we call it the what we don't call it the Shannon diversity index is what it is called this is an

502
00:58:47,000 --> 00:58:54,000
ecologically inspired term. It is just Shannon entropy. For those of you who are not familiar with Shannon and entropy.

503
00:58:54,000 --> 00:59:00,000
I will briefly describe it. So on the left you see something with very low entropy, very low diversity.

504
00:59:00,000 --> 00:59:10,000
You kind of know what you're getting with that thing where it fits, what context it's used in, and then over on the very far right you can you would have no idea when and where you might see this thing.

505
00:59:10,000 --> 00:59:14,000
So it would be, you might consider it to be very diverse.

506
00:59:14,000 --> 00:59:19,000
I mean, let me just take you know I mentioned the 8 different contexts a few slide previously.

507
00:59:19,000 --> 00:59:31,000
That's what you know. you have 8 of these diversity measures for each hashtag, and then you just multiply them by some waiting, and you come up with this ensemble diversity index and that value would be an

508
00:59:31,000 --> 00:59:39,000
indication of how general that hashtag is the higher the value the more general it is, the lower the value, the less general it is, the more specific.

509
00:59:39,000 --> 00:59:42,000
So i'll i'll get to the data the all important data.

510
00:59:42,000 --> 00:59:47,000
So you know, we want to archive for it we we used Twitter's 1% spritzer stream.

511
00:59:47,000 --> 00:59:55,000
And we got 52 months of data between October the twentieth 16 in December the twentieth, 2146,000,000 English language tweets.

512
00:59:55,000 --> 01:00:00,000
We have many more than that other languages, and 360,000 hashtags.

513
01:00:00,000 --> 01:00:06,000
So we took all that data. We boil it down into a hashtag network the co-occurring hashtags.

514
01:00:06,000 --> 01:00:11,000
And then we used this network to calculate a lot of our measures.

515
01:00:11,000 --> 01:00:14,000
So here's some of the hierarchy so you know.

516
01:00:14,000 --> 01:00:22,000
Here's the data community, or shall we say that the top 10 most diverse hashtags within the data community.

517
01:00:22,000 --> 01:00:28,000
There are about over 2,000 other hashtags in this community alone we can see at the top.

518
01:00:28,000 --> 01:00:37,000
You have pretty familiar words there that you may recognize and then if we look across other communities of hashtags.

519
01:00:37,000 --> 01:00:49,000
So here's the there's the data community the beer community, the coffee community, and then the dog community at the time, you know, the most diverse, the most general hashtags are very familiar terms You would recognize most all

520
01:00:49,000 --> 01:00:53,000
of them, and then at the bottom, the least diverse hashtags the most specific, you know.

521
01:00:53,000 --> 01:00:58,000
I I don't i've never seen these before. they they look like they sort of make sense in the community.

522
01:00:58,000 --> 01:01:02,000
But they're they're not they're very narrowly used, and

523
01:01:02,000 --> 01:01:07,000
There's a beautiful tail here, which is based on our our findings.

524
01:01:07,000 --> 01:01:16,000
Hashtag love happens to be the most diverse hashtag, according to our method. according to the archive org data that we have.

525
01:01:16,000 --> 01:01:21,000
So that's that's a beautiful story there and you can see some of the related communities.

526
01:01:21,000 --> 01:01:34,000
To the love community. These are the most tightly corresponding communities, and you know, not shown our thousands of other communities and hundreds of thousands of other hashtags.

527
01:01:34,000 --> 01:01:38,000
But you can. You can certainly, you know, dive into it.

528
01:01:38,000 --> 01:01:41,000
So what can be used it? It can be used as a hashtag recommended.

529
01:01:41,000 --> 01:01:44,000
So if hashtag donation maybe also hashtag dog.

530
01:01:44,000 --> 01:01:51,000
But the idea of this general to specific ordering may I suggest that you don't want to do a hashtag dog.

531
01:01:51,000 --> 01:01:53,000
Maybe also hashtag donation that doesn't really work

532
01:01:53,000 --> 01:01:57,000
You can look at hierarchical hashtag topic models.

533
01:01:57,000 --> 01:02:00,000
So we know that hashtag ai is part of the data community.

534
01:02:00,000 --> 01:02:04,000
So that's you know. good information no we also know that it is the data community.

535
01:02:04,000 --> 01:02:08,000
We can do social science investigations. What does hashtag Minnesota mean?

536
01:02:08,000 --> 01:02:20,000
And when was that meant? It certainly was more geographical prior to May of 2,020, with the murder of George Floyd in Minneapolis, when it Minnesota, was now associated with the Black lives

537
01:02:20,000 --> 01:02:23,000
matter movement, and you, you know, can use other platforms to.

538
01:02:23,000 --> 01:02:29,000
We showed Twitter. But you know this. any any tags will will do for this.

539
01:02:29,000 --> 01:02:33,000
So. thank you for your time. I appreciate it and thank you to the presenters again.

540
01:02:33,000 --> 01:02:40,000
And if you're interested these are the names of the papers relevant, Thank you so much, Spencer.

541
01:02:40,000 --> 01:02:48,000
We have some questions that are coming in, and if you and the audience have additional questions, please do drop them into the Q. and A.

542
01:02:48,000 --> 01:02:54,000
The one thing that I was just struck by that data. Beer, coffee dog, hashtags are life.

543
01:02:54,000 --> 01:02:58,000
So hold tight there for a second before we go on.

544
01:02:58,000 --> 01:03:11,000
I wanna acknowledge i'm remiss in not acknowledging art rhino who is in the audience today that that fantastic video that we just watched aren't artists here watching along with us So Thank

545
01:03:11,000 --> 01:03:21,000
you art for for that submission and for your creative work up next in our show. is another video, and another presenter who is also in the audience.

546
01:03:21,000 --> 01:03:29,000
So we're gonna hear from Jim salman who is going to tell us about his Internet archive enabled journey as a digital humanities.

547
01:03:29,000 --> 01:03:43,000
Citizen science. So caitlin let's roll Hi I'm Jim Salmons and this lightning talk is a whirlwind trip through my post-cancer journey of rebirth as a digital humanity citizen scientist

548
01:03:43,000 --> 01:03:48,000
at how the inspiration of an access to the Internet Archive made that possible.

549
01:03:48,000 --> 01:04:00,000
Starting in 2,012, through 2,014. Both my wife, Tim Labbitsky and I had terrifying cancer battles that we fortunately survived to pay it forward in celebration of our 20 fifth wedding

550
01:04:00,000 --> 01:04:08,000
anniversary. we funded the digitization of the 48 issues of the Apple computer focus soft Talk magazine into the Internet.

551
01:04:08,000 --> 01:04:14,000
Archive, soft talk has a special place in my heart. As I was a reader.

552
01:04:14,000 --> 01:04:28,000
Advertiser writer, and during the time of its explosive growth and executive at softball publishing, where I designed and helped develop the software that ran the back office production and advertising processes as part of our

553
01:04:28,000 --> 01:04:40,000
initial soft talk preservation project. Tim and I went to the midwest scanning center of the Archive, where we saw firsthand and participated in the amazing behind-the-scenes scanning service It

554
01:04:40,000 --> 01:04:49,000
puts digital collections into the Internet Archive, not wanting our contribution to end with simply getting the digital edition of soft talk into the archive.

555
01:04:49,000 --> 01:04:55,000
We followed up with activity. That was the true beginning of my rebirth as a digital humanity.

556
01:04:55,000 --> 01:05:09,000
Citizen scientist, as I learned about the challenges of text and data mining of digital collections within the cultural heritage domain, I decided to create a ground. true storage format would support an integrated model of a

557
01:05:09,000 --> 01:05:26,000
Magazine's complex document structures and content depiction between 2015, or 19, I developed a personal learning network of largely Eu and Uk based mentors and collaborators to support the development of the Magazine Gts

558
01:05:26,000 --> 01:05:37,000
metadata format based on international museum, ontology standards with posters and papers, accepted the Eu-based digitization conferences and workshops.

559
01:05:37,000 --> 01:05:45,000
My work, initially focused on the structure and content of advertisements and soft talk.

560
01:05:45,000 --> 01:05:54,000
Everything came to crashing halt when I suffered a devastating spinal cord injury in 2,020 during the year and a half of my rehab and recovery.

561
01:05:54,000 --> 01:06:07,000
The digital humanities domain saw explosive growth in the use of machine learning technologies. As I reinvigorate my research, I have expanded my focus from the softball collection.

562
01:06:07,000 --> 01:06:20,000
Consider the challenges of investigating the massive Internet Archive collection of computer magazines, consisting of tens of thousands of issues of publications, dozens of languages published all around the world.

563
01:06:20,000 --> 01:06:26,000
Looking at the impact of computers and emergence of the digital world we live in today.

564
01:06:26,000 --> 01:06:33,000
My initial exploration of this expanded collection is focused on development of a ground truth.

565
01:06:33,000 --> 01:06:37,000
Data set of computer magazine talk or table of contents.

566
01:06:37,000 --> 01:06:45,000
Pages. talks serve as a seducco, puzzle-like set of hints about the document structures of a magazine.

567
01:06:45,000 --> 01:06:56,000
This data set will be invaluable for training machine learning models to help move digitization pipelines from within page to whole document, layout recognition.

568
01:06:56,000 --> 01:07:04,000
My goal now is to stimulate collaboration between my Eu and Ukraine friends with new partners from Stanford libraries.

569
01:07:04,000 --> 01:07:17,000
It's Ai lab the computer history, museum and the Internet archive, the forge, a research consortium to further preserve and make accessible for scholarly research and public interest.

570
01:07:17,000 --> 01:07:23,000
The computer magazines met a collection at the Internet Archive on behalf of myself and Tim Lubbitsky.

571
01:07:23,000 --> 01:07:40,000
Thank you to the archive and webinar organizers for inviting me to present this lightning talk. and thanks to you, Jim, for sharing your inspiring story, your video, your work, and for sponsoring the digitization of the

572
01:07:40,000 --> 01:07:44,000
materials that are now available to all at the Internet Archive.

573
01:07:44,000 --> 01:07:52,000
That your table of contents, Tooc, or talk work is of hi interest to us.

574
01:07:52,000 --> 01:08:01,000
I know also, from my previous work with the biodiversity Heritage Library, that those tables of contents are really that's where it's all at in terms of the structural metadata for

575
01:08:01,000 --> 01:08:12,000
the for the article So We'll be following up with learn a little bit more about what you're doing there. So up next i'd like to welcome Emmauel tranos to the screen.

576
01:08:12,000 --> 01:08:17,000
Who's gonna tell us about the relationship that exists between the Web and cities?

577
01:08:17,000 --> 01:08:22,000
This is a really interesting talk. I know You're gonna love it so over to you.

578
01:08:22,000 --> 01:08:27,000
Emmanuel Kurtis. Thank you so much for this.

579
01:08:27,000 --> 01:08:35,000
Let me try my screen, and also tell you how excited about him to be part of this very cool A.

580
01:08:35,000 --> 01:08:44,000
So. my name is amado janos i'm a reader in quantitative. He wants geography at the University of Bristol and the islands.

581
01:08:44,000 --> 01:08:52,000
You include in the Uk. and I guess i'm i'm of this rare breed of geographers who have a vaccine with the with the Internet.

582
01:08:52,000 --> 01:09:07,000
So today, i'm gonna give you a brief overview of our research using data from the Internet archive to understand the link between web and seeds, we're using such data to understand every Internet.

583
01:09:07,000 --> 01:09:13,000
But also other interesting a it's your account so what data. do we use?

584
01:09:13,000 --> 01:09:21,000
We use the We're using data that we say curated by the British library here in Uk.

585
01:09:21,000 --> 01:09:26,000
This is data called the Disc Uk Web Domain data set.

586
01:09:26,000 --> 01:09:40,000
And this is simply all. This contains all the archived web pages from data Archive under the Dot Uk Top level domain during the 996 to 2,012 P.

587
01:09:40,000 --> 01:10:01,000
The business library did something very clever. here. they scanned the web text of all these archived web pages and created a different further subset, which only includes these archive web page which contain Hey Uk postcode.

588
01:10:01,000 --> 01:10:07,000
We feed their web text, and I believe you can see the link to these data sets in the top.

589
01:10:07,000 --> 01:10:14,000
So we started our research with almost half a 1 billion of of lines which look like this.

590
01:10:14,000 --> 01:10:29,000
We know that this archive Url is contains within its work Text: this Uk Post code, and this post code refers to a very small and area almost a block in the Vj.

591
01:10:29,000 --> 01:10:33,000
And you also know the timeest of these, a a archive.

592
01:10:33,000 --> 01:10:38,000
On September of May, the ninth of September, of 2,008.

593
01:10:38,000 --> 01:10:51,000
So what do we do with these data? first thing we use these data to create a a measure of the online content of local interest.

594
01:10:51,000 --> 01:11:08,000
And we must. These these data with that and a large individual survey. And we're able to illustrate that we are going ability of online content, of local interest actually attracts individuals online.

595
01:11:08,000 --> 01:11:15,000
We knew a lot about the fact that that's not pushing the videos, you know, to come back to their Internet to spend more time online.

596
01:11:15,000 --> 01:11:25,000
But this was the first time we were able to say something about the other Maxwell factors that are attracting me with us to spend more time online.

597
01:11:25,000 --> 01:11:33,000
And importantly, to do this at the very local state at the difference scale at a different a A.

598
01:11:33,000 --> 01:11:40,000
A. I started we used such data in order to understand economic classes.

599
01:11:40,000 --> 01:11:46,000
And when I say economic clusters, I refer to these neighborhoods within cities.

600
01:11:46,000 --> 01:11:54,000
That host, very, You know, specific economic activities, these very specialized a neighborhoods or specific.

601
01:11:54,000 --> 01:12:15,000
We focused on solving. This is a very well-known tech cluster here in London, and we used archived web data in order to map the evolution of this economic cluster over space and time but importantly to map the

602
01:12:15,000 --> 01:12:21,000
evolution of this cluster also in terms of the types of economic activities that took place.

603
01:12:21,000 --> 01:12:35,000
You know, within this cluster and we're. able you know to extract meaningful types of economic activities, much more detailed than official data would have enabled us to do so.

604
01:12:35,000 --> 01:12:47,000
Ideally We're Chasing. the scale of of analysis, and we're moving, you know, from a small neighborhood in in London to the whole of the Uk we utilize.

605
01:12:47,000 --> 01:12:56,000
We employed such archives and web data in order to test the economic effects.

606
01:12:56,000 --> 01:13:03,000
That's the early adoption of web technologies can generate to reaches here in the Uk.

607
01:13:03,000 --> 01:13:15,000
So we're able to build you know measures of the volume of online commercial, you know, content buck from 2,000.

608
01:13:15,000 --> 01:13:24,000
And we're able to to link you know these measures from 2,000 to specific regions within the Dvd.

609
01:13:24,000 --> 01:13:33,000
We then utilized economics techniques and we're able to illustrate an interesting license the volume of online content.

610
01:13:33,000 --> 01:13:45,000
But from 2,000, which to us represents the area. Adoption of web technologies is actually associated with positive economic productivity effects.

611
01:13:45,000 --> 01:13:56,000
But these productivity effects a long lasting. They are long term, you know, a a positive effects that regions work,

612
01:13:56,000 --> 01:14:16,000
Who Who? who employed this technology, Ariel, are able to enjoy it for longer, for quite lengthy a a time periods. and to last, but not least at the similar state we used such archived web data and more specifically

613
01:14:16,000 --> 01:14:27,000
html links between commercial websites to predict trade between different Uj reasons.

614
01:14:27,000 --> 01:14:38,000
Watch the Cdcally, we're able to make out of sample predictions, using my singular link algorithms regarding this regional trade flow.

615
01:14:38,000 --> 01:14:46,000
Was this important because there is hardly any data for trade between retail so it's not the small state.

616
01:14:46,000 --> 01:15:00,000
So by using this freely available, you know data that data like I, if you know, collects for all of us, were able to make predictions for such an important policy element and have local authorities utilize.

617
01:15:00,000 --> 01:15:18,000
These data that we generated. So all you know, using this filler variable data, were able to map the evolution and the geography of the engagement with the Internet, especially at this early status and trust me there is hardly

618
01:15:18,000 --> 01:15:26,000
any other data, you know, that can go so back in time and also so granular by doing that, we're able to draw important lessons.

619
01:15:26,000 --> 01:15:43,000
Righting the deployment of other future technologies, and also what able to understand economic activities at the very day level, both in terms of space and time, but also in kind of context, take place.

620
01:15:43,000 --> 01:15:56,000
Within. And so, between scenes all of these research can be found in in in various papers who have published in publications as well, that you can find on my website.

621
01:15:56,000 --> 01:16:04,000
Again. Thank you so much for having me here today. Thank you, Emmanuel, for sharing your research.

622
01:16:04,000 --> 01:16:07,000
Oh, we do have a question for you and for Spencer.

623
01:16:07,000 --> 01:16:12,000
But we wanna wrap up today with a final video from from our session.

624
01:16:12,000 --> 01:16:17,000
So we're gonna welcome Tom galley back to the screen virtually.

625
01:16:17,000 --> 01:16:29,000
And he's gonna share his his research and his work into the forgotten novels of the nineteenth century

626
01:16:29,000 --> 01:16:34,000
I started reading old books back in the 1,900 sixtys when I was still a child.

627
01:16:34,000 --> 01:16:42,000
So how did I decide what books to read? Well, we had a lot of books at home, and I read some of those.

628
01:16:42,000 --> 01:16:53,000
There was a small public library nearby, and I like to browse the shelves there, too, and when I got my allowance I would go to the local bookstore.

629
01:16:53,000 --> 01:16:59,000
Look through those shelves and maybe buy a couple of paperbacks that got my eye.

630
01:16:59,000 --> 01:17:05,000
The books I happen to see on those shelves would shape my reading choices in the years to come.

631
01:17:05,000 --> 01:17:14,000
When I started reading nineteenth century novels, I naturally gravitated towards authors and books that I had seen on those shelves.

632
01:17:14,000 --> 01:17:18,000
Charles Dickens Jane Austen Nathaniel Hawthorne.

633
01:17:18,000 --> 01:17:23,000
Crime and punishment, the adventures of Huckleberry Finn.

634
01:17:23,000 --> 01:17:37,000
In other words, the classics. It was only later, after I had access to a large university library that I discovered the vast number of other novels published in the nineteenth century.

635
01:17:37,000 --> 01:17:43,000
Their pages were yellow and brittle, and most had never been reprinted.

636
01:17:43,000 --> 01:17:48,000
But after I graduated from college I could no longer access those books.

637
01:17:48,000 --> 01:18:07,000
I was stuck with the classics again. so fast forward to around the year, 2,010 libraries around the world were now scanning their old books, and the Internet Archive was making those books available online for free now anyone in the

638
01:18:07,000 --> 01:18:17,000
world could read all of those old novels, including the thousands that publishers had not reprinted and marketed as classics.

639
01:18:17,000 --> 01:18:24,000
So in 2,021 just for fun, compile the list of nineteenth century novels at the Internet.

640
01:18:24,000 --> 01:18:30,000
Archive chose only novels that nobody seemed to be reading anymore.

641
01:18:30,000 --> 01:18:36,000
But the once popular fiction that had been overlooked by the classics industry.

642
01:18:36,000 --> 01:18:44,000
You can find that list on the Internet Archives blog under the title Forgotten Novels of the Nineteenth Century.

643
01:18:44,000 --> 01:18:53,000
I've enjoyed dipping into those books maybe you will too

644
01:18:53,000 --> 01:19:06,000
So we'll share that link out to the to tom's blog for everyone in the email follow up. and I'm sure you'll find some new things that you haven't read for a while thanks Duncan for for

645
01:19:06,000 --> 01:19:15,000
sharing that out. I would like to bring spencer and Emmanuel back to the screen for a couple of questions, and the first one's dispenser.

646
01:19:15,000 --> 01:19:22,000
It's have there been changes in how twitter users use hashtags over time, especially more recently.

647
01:19:22,000 --> 01:19:28,000
That's a really good question. we haven't looked into it in that respect yet. but that is on the docket.

648
01:19:28,000 --> 01:19:30,000
So you know again i'll reference the hashtag Minnesota.

649
01:19:30,000 --> 01:19:37,000
Example, where, prior to May, the twentieth 20 all geography, I mean, it has to do with States capitals.

650
01:19:37,000 --> 01:19:43,000
Right, no mention of civil rights or anything. And then, after may the twentieth 20 very much non geographical.

651
01:19:43,000 --> 01:19:48,000
Well, incidentally geographical, having to do with black lives matter and the murder of George Floyd.

652
01:19:48,000 --> 01:19:53,000
So that that's an example. But it's a great question and something we intend to dive into.

653
01:19:53,000 --> 01:19:59,000
Thank you for that. and a a follow up question for Emmanuel.

654
01:19:59,000 --> 01:20:03,000
Is your research methodology extensible to other geographic areas?

655
01:20:03,000 --> 01:20:07,000
Or would you need to change your approach for studying areas outside the Uk?

656
01:20:07,000 --> 01:20:13,000
No, it is absolutely extendable and replicable.

657
01:20:13,000 --> 01:20:28,000
Outside of the Uk the main differences having you know, available, you know, datasets or subsets, you know, from payment that guy. and actually this one of the value of these data, because it's these are openly up available data source

658
01:20:28,000 --> 01:20:39,000
sources, and by using these data, what actually i'll perform some of the official data sources that's that's fascinating.

659
01:20:39,000 --> 01:20:50,000
I I know that there are some additional questions but i'm also sensitive about our time. So what i'm going to ask everyone to do is we're we're gonna have everyone has shared their contact.

660
01:20:50,000 --> 01:21:00,000
Information. So if you have additional questions for any of our presenters today, maybe we can follow up offline or after the session.

661
01:21:00,000 --> 01:21:08,000
So thank you, Spencer Emmanuel, for for your conversations today, and for taking some time to field some questions.

662
01:21:08,000 --> 01:21:17,000
Thank you much. and also, thanks to to Jim, and to Tom for the videos that they offered up Jim and Timeline as well.

663
01:21:17,000 --> 01:21:22,000
So here we are at the end we've made it

664
01:21:22,000 --> 01:21:25,000
I I wanna give it an acknowledgement to the

665
01:21:25,000 --> 01:21:38,000
Some people who helped guide this series from its start from inception, and we pull together an advisory group for our series, And that included some real heavy headers in in the digital humanity space.

666
01:21:38,000 --> 01:21:52,000
So a big Thank you. to Dan Cohen, from Northeastern to Mickeya Foster, from Broward County Library, Mike Furlough from Hoti Trust, and from Harriet Green at Washington University, in St.

667
01:21:52,000 --> 01:22:06,000
Louis. they've they really helped shape the talks that that we saw here today and help give guidance on how we should frame a conversation along a long tail Conversation a longitudinal chat If you will among our digital

668
01:22:06,000 --> 01:22:10,000
humanity scholars. So, thanks to the advisory committee for

669
01:22:10,000 --> 01:22:15,000
For helping us out with that i'm gonna go a little off script.

670
01:22:15,000 --> 01:22:27,000
I don't know if Brewster is still on the call I think that he is but I I would like to offer Brewster a chance if he's still available and he's near his computer to come on screen and

671
01:22:27,000 --> 01:22:31,000
give us a little a wrap up you know.

672
01:22:31,000 --> 01:22:34,000
What What do you think of Everything that you've seen here today?

673
01:22:34,000 --> 01:22:39,000
And across the series. Oh, this is so inspiring and so fun!

674
01:22:39,000 --> 01:22:45,000
And it's just to be able to see that the light and humorous nature of this.

675
01:22:45,000 --> 01:22:48,000
But as well as you know, answering these real questions.

676
01:22:48,000 --> 01:22:55,000
I, you know, and I, Of course I spring to the We have this more to to to just share with you.

677
01:22:55,000 --> 01:23:01,000
We have periodicals that have that electronic data processing schools is schools.

678
01:23:01,000 --> 01:23:10,000
Anyway. i'm so glad that this is going on I'm. and I think they only need format at least works for me.

679
01:23:10,000 --> 01:23:14,000
So thank you very much for pulling this together, Chris and everyone.

680
01:23:14,000 --> 01:23:18,000
Yeah, thanks. I I think this the lightning format it's nice to have the variety right.

681
01:23:18,000 --> 01:23:30,000
You can do some deep dives on some individuals topics and Then, get a broad overview of the wealth of materials and the wealth of research that's happening to in and around the collections at the Internet.

682
01:23:30,000 --> 01:23:36,000
Archive. What I would say is, we wind down here today is this is the start of a conversation.

683
01:23:36,000 --> 01:23:48,000
Let's think of this as the start of a conversation not as the end of one with this series part of what we wanted to do in bringing this together was to one raise awareness just that there are digital humanities projects that are actively, using

684
01:23:48,000 --> 01:23:51,000
the collections and the infrastructure at the Internet Archive.

685
01:23:51,000 --> 01:23:55,000
And so what we were hoping was to bring some of those people together.

686
01:23:55,000 --> 01:24:01,000
Provide, some visibility into the work that everyone is doing, and start that conversation, so that we can do more together.

687
01:24:01,000 --> 01:24:21,000
So to wind down here today on Behalf of the Internet archive and our present presenters today i'd like to thank you all for your time and your participation and a big thanks to everyone who's joined in the multiple sessions that we've